Four of these factors are essential for cleavage in vitro at all poly(A) sites so far examined. Cleavage between these two elements is usually on the 3′ side of an A residue ( 11) and, in vitro, is mediated by a large, multicomponent protein complex that can be separated into five distinguishable factors ( 14, 65). Typically, an almost invariant AAUAAA hexamer lies 20 to 50 nucleotides (nt) upstream of a more variable element rich in U or GU residues ( 11, 23, 48). The core poly(A) signal in vertebrates consists of two recognition elements flanking a cleavage-polyadenylation site. Furthermore, we show that the strong simian virus 40 (SV40) late poly(A) site assembles considerably more rapidly in vivo than other sites known to be weaker.

We report here that the assembly of the cleavage-polyadenylation apparatus in vivo is more complex than expected, being a gradual multistep process. Until now, however, there has been no information on the nature of the process in vivo. In contrast, no natural assembly intermediates have been observed for the cleavage-polyadenylation apparatus in vitro with crude extracts ( 38).

Spliceosome assembly occurs naturally in discrete steps ( 61) both in vivo ( 7) and in crude extracts in vitro ( 22). The basic reaction in vertebrates involves only one site on the RNA, not three as for splicing, and the core apparatus is comprised entirely of proteins-no small nuclear RNAs are required ( 14, 65). The SV40 late poly(A) site, one of the strongest, assembles several times faster than the weaker SV40 early or synthetic poly(A) site.Ĭompared to splicing, the cleavage-polyadenylation process is expected to be fairly straightforward. We compared strong and weak poly(A) sites. The simplest explanation for this target size effect is that the assembly process progressively sequesters more and more of the RNA surrounding the poly(A) signal up to a maximum of about 200 nucleotides, which we infer to be the domain of the mature apparatus. This indicates that a brief period of assembly is sufficient for the poly(A) signal to shield itself from a short (50- to 70-nucleotide) antisense sequence but that more assembly time is required for the signal to become immune to the longer ones (∼200 nucleotides). Relief from inhibition occurred earlier for shorter antisense sequences than for longer ones. Based on the known rate of transcription, we estimate that the cleavage-polyadenylation process takes between 10 and 20 s for the SV40 early poly(A) site to complete in vivo. Antisense inhibition was unaffected when the inverted signal was moved upstream. The antisense inhibition was gradually relieved when the inverted signal was moved increasing distances downstream, presumably because cleavage and polyadenylation occur before the polymerase reaches the antisense sequence. An inverted copy of the poly(A) signal placed immediately downstream of the authentic one inhibited processing by means of sense-antisense duplex formation in the RNA. We have devised a cis-antisense rescue assay of cleavage and polyadenylation to determine how long it takes the simian virus 40 (SV40) early poly(A) signal to commit itself to processing in vivo.

Sv40 poly a snapgene